Prevalence and Significance of Impaired Microvascular Tissue Reperfusion Despite Macrovascular Angiographic Reperfusion (No-Reflow)

Background The relevance of impaired microvascular tissue-level reperfusion despite complete upstream macrovascular angiographic reperfusion (no-reflow) in human stroke remains controversial. We investigated the prevalence and clinical-radiological features of this phenomenon, and its associations with outcomes in three international randomized controlled thrombectomy trials with pre-specified follow-up perfusion imaging. Methods In a pooled analysis of the EXTEND-IA ( ClinicalTrials.gov number NCT01492725 ), EXTEND-IA TNK ( NCT02388061 ) and EXTEND-IA TNK Part-two ( NCT03340493 ) trials, patients undergoing thrombectomy with final angiographic extended Thrombolysis In Cerebral Ischemia 2c-3 score for anterior circulation large vessel occlusion and 24-hour follow-up CT or MRI perfusion imaging were included. No-reflow was defined as regions of visually demonstrable persistent hypoperfusion on relative Cerebral Blood Volume or Flow maps within the infarct and verified quantitatively by >15% asymmetry compared to a mirror homologue in the absence of carotid stenosis or re-occlusion. Results Regions of no-reflow were identified in 33 of 130 patients (25.3%), encompassed a median of 60.2% (Interquartile range 47.8-70.7%) of the infarct volume, and involved both subcortical (n=26/33,78.8%) and cortical (n=10/33,30.3%) regions. Patients with no-reflow had a median 25.2% ([Interquartile range 16.4-32.2%],p<0.00001) relative Cerebral Blood Volume interside reduction and 19.1% (Interquartile range 3.9-28.3%,p=0.00011) relative Cerebral Blood Flow reduction but similar mean-transit-time (median -3.3%, Interquartile range -11.9-24.4%,p=0.24) within the infarcted region. Baseline characteristics were similar between patients with and without no-reflow. The presence of no-reflow was associated with hemorrhagic transformation (aOR=1.79,95%CI2.32-15.57,p=0.0002), greater infarct growth (ß=11.00,95%CI5.22-16.78,p=0.00027), reduced National Institutes of Health Stroke Score improvement at 24-hours (ß=-4.06,95%CI-6.78--1.34,p=0.004) and being dependent or dead at 90-day as assessed on the modified Rankin Scale (aOR=3.72,95%CI1.35-10.20,p=0.011) in multivariable analysis. Conclusion Cerebral no-reflow in humans is common, can be detected by its characteristic perfusion imaging profile using readily available sequences in the clinical setting, and is associated with post-treatment complications and being dependent or dead. Further studies evaluating the role of no-reflow in secondary injury after angiographic reperfusion are warranted. Classification of evidence This study provides Class II evidence that cerebral no-reflow on CT/MRI perfusion imaging at 24-hours is associated with post-treatment complications and poor 3-month functional outcome.